Text-to-Speech Studio (beta)

Prompt formatting

• Put each speaker on its own line using a tag like <Voice 1>: line of dialogue.
• The label inside <…> can be any name or number, but keep the closing > immediately followed by a colon and a space.
• When you rely on a single voice sample, you may skip tags and submit plain text.
• Avoid placing tags mid-sentence — make sure <Voice …> appears right after a line break or leading whitespace.
• Skip manual typing: click the + button next to any voice card and the tag will be inserted for you.
• Pause tags are supported as well: [pause] and [pause:ms] (milliseconds), but they are still finicky, so double-check the rendered audio.

Examples

<Voice 1>: Welcome to our showcase.
<Voice 2>: I will walk you through the feature set.

Single voice only — no tags required, just enter the full script as plain text.